HOW TO THINK ABOUT THE EVOLUTION OF THE RATIO OF Mg TO Ca IN SEAWATER

Measurements on fluid inclusions (Timofeeff and others, 2006) indicate that over the last 110 million years, the Mg to Ca ratio in seawater has risen by at least a factor of four. Those by Zachos and others (2003) on planktic foraminifera suggest that about half of this rise occurred during the last 55 million years. Although Coggon and others' (2010) measurements on ridge-flank basalts are consistent with a rise of this magnitude, they indicate that it occurred after 46 million years ago and perhaps even as recently as the last 15 million years. Based on the Mg to K trend in 36 million-year-old halite-hosted fluid inclusions, Brennan and others (2013) confirm Zimmerman's (2000) conclusion that the Mg content of sea salt was about 35 percent lower than now at that time. Hence, if the Coggon and others' (2010) results are to be believed, then the rise in Mg must have been accompanied by a drop in calcium. When account is taken that the CO2 content of the atmosphere 35 million years ago was about five times larger than it is today, then perhaps at that time, the dissolution rate of limestone was greater than today's. If so, then the large increase in Mg to Ca ratio could be explained by a combination of a decrease in sea level which slowed the rate of dolomite formation and a drop in atmospheric CO2 content which slowed the delivery by rivers to the sea of Ca derived from limestone weathering. Unfortunately, as the deposition of gypsum prior to halite formation removed virtually all the calcium from the brine pools in which the halite subsequently formed, there is no record of Ca in Cenozoic-age fluid inclusions. Hence, in order to resolve the difference between the large Mg/Ca rise over the last 40 million years suggested by ridge-plank calcites and, the smaller one suggested by planktic foraminifera, an additional proxy for Ca is needed. As I suggested many years ago (Broecker, 1970), one such candidate is the uranium content of corals.